Not a moan, but constructive info for Fred.
I now have about a dozen apps written in PB. Yesterday I decided to upgrade to PB3.72. Minor differences and a bug (reported) meant that it took me 5 hours to get these programs working as they were before.
Multiply those 5 hours by the size of the user base and you have a frightening amount of effort associated with a new version.
Of course, new versions are needed to add the features we all want and to fix bugs. However, the calculation above suggests to me that less frequent upgrades and even more checking may be popular.
NOT (moan)
A new version can cause all kinds of trouble. Personally I do not recompile every application as soon as there is a new PB version available.
If Fred releases frequent new versions, you still have the option not to download it. If you download every second or third new release, you can pretend that the other ones did not exist. It will save you some trouble and people like me can go on enjoying frequent releases.
Then again, I could be wrong... 8O
/Gantry
If Fred releases frequent new versions, you still have the option not to download it. If you download every second or third new release, you can pretend that the other ones did not exist. It will save you some trouble and people like me can go on enjoying frequent releases.
Then again, I could be wrong... 8O
/Gantry
*** In the end we will remember not the words of our enemies, but the silence of our friends. (Martin Luther King Jr.) ***
I do not see how less frequent updates would solve your problem.
If only every second update would be released, the changes for each
update would be twice as much, and as a result of that, the amount of
work you'd have to to would double.
My point is: Updates bring changes, changes bring work, the only question
is: do you want them all at a time, or in small pieces?
> [...] and even more checking may be popular.
Hmm, we run into a pretty big problem here. Who whould do that?
I mean, we are allready doing *public beta testing*. How do you want to
improove that?
Timo
If only every second update would be released, the changes for each
update would be twice as much, and as a result of that, the amount of
work you'd have to to would double.
My point is: Updates bring changes, changes bring work, the only question
is: do you want them all at a time, or in small pieces?
> [...] and even more checking may be popular.
Hmm, we run into a pretty big problem here. Who whould do that?
I mean, we are allready doing *public beta testing*. How do you want to
improove that?
Timo
quidquid Latine dictum sit altum videtur
Thanks Freak and Gantry for your sensible comments. I agree with what you say, maybe your best advice is not to download every release.
However, more automatic checking by Fred would be welcome. He has explained that he uses extensive software checks before every release but I think these checks need improving.
For example I found this bug in 3.72
a.l=2
b.f=1
If a<b: Debug "error": EndIf
It is hard to believe that errors this simple can't be found by simple checking software. The core language elements like expression evaluation and comparison could be tested by a simple PB program.
However, more automatic checking by Fred would be welcome. He has explained that he uses extensive software checks before every release but I think these checks need improving.
For example I found this bug in 3.72
a.l=2
b.f=1
If a<b: Debug "error": EndIf
It is hard to believe that errors this simple can't be found by simple checking software. The core language elements like expression evaluation and comparison could be tested by a simple PB program.